1. Field of the Invention
The present invention relates to an image processing method, and more particularly, to a method that utilizes image categorizing techniques for image de-interlacing.
2. Description of the Prior Art
Please refer to FIG. 1. FIG. 1 is a system block diagram of a conventional de-interlacing apparatus 100. The de-interlacing apparatus 100 includes an edge detection apparatus 110 and an interpolation circuit 120. The edge detection apparatus 110 receives an input image P_in (for example, the input image P_in can be a field of a video data, wherein the field corresponds to a frame, and the input image P_in only includes a portion of scan lines of the frame), and then performs an edge detection operation on the input image P_in to generate edge information leg corresponding to the input image P_in. Next, the back stage interpolation circuit 120 performs an interpolation on the input image P_in according to the edge information I_eg, in order to generate a portion of missing scan lines in the image P_in, so as to generate an output image P_out corresponding to the input image P_in (the output image P_out includes the portion of missing scan lines in the input image P_in). However, the conventional method does not always accurately perform the interpolation on the input image according to the edge information of the image, and therefore the de-interlacing results are not ideal.